[tomp]

Really interesting reasoning. I'm almost tempted to say "extremely unscientific".

1. "no evidence of microplastics being harmful" sounds very much like WHO's "no evidence of human transmission". Personally, I'd say that in both scenarios, the outcome we should be most worried about is tail risks - what if we suddenly realize the situation is much worse than we imagined, but it's already too late...

2. Is "climate change" so bad? I mean, would it be really so bad if the Earth warmed up a bit, and all of Siberia, Greenland and the Antartic became livable (and arable) land? More food, more space for new humans, ... all good! Except, obviously, 2 things - biodiversity (which you want to ignore) and tail risks - what if the Earth warms up too much, and we "trigger" some unexpected process (e.g. thawing of permafrost releases vast quantities of extremely potent greenhouse gases).

I think most people are simply really confused about their reasoning process. Microplastics? Ignore tail risks. Global warming? Tail risks! GMOs? Ignore tail risks. 5G? Tail risks! Vaccines? Ignore tail risks. Endocrine disruptors? Tail risks! (Of course, the exact configuration of positions is highly correlated with one's political orientation.)

[pietrovismara]

2. Of course it's incredibly bad. It will displace entire populations and alter permanently entire ecosystems. The effects will be cascading all over the world and we have no idea what the consequences will be in the long term.

It is definitely a very concerning prospective.

[rorykoehler]

You've articulated perfectly something I've been struggling with on conversations here on hackernews in your last paragraph. People can be incredibly short termist when it suits their agenda but some things are dangerously irreversible if we get them wrong. I liken it to a child dismantling a toy to see how it works and then not being able to put it together again but on a civilisation impacting level.

[TeMPOraL]

I'd argue the two basic heuristics I apply in this and similar cases are pretty scientific:

1) Strong effects are obvious in the data; conversely, if you have to dig deep and play with statistical significance to prove there's a connection, the effect is very weak - it's a scientific curiosity, sure, but "to zeroth order of approximation" you may as well say it doesn't exit.

2) When ascertaining the possible impact of something in the future, your estimate will be more accurate if you have a clue about mechanisms of action involved.

Microplastics seem to fail both heuristics at the moment - there isn't any obvious impact that we've measured, and there isn't any obvious mechanism of action that could imply serious danger. There most likely are issues caused by them - however, we've been living with them for decades now, so if this was an emergency-level threat, we'd probably see something by now.

Compare that with worrying whether coffee/red meat/artificial sweetener/vaccine/5G will give you cancer - maybe it will, maybe it won't, but if the chance of it was remotely worth worrying about, it would be blindingly obvious in the data by now. Contrast that with "no evidence of human transmission" - not enough data to check heuristic 1, fails heuristic 2 due to similarity to other coronaviruses with known mechanisms of human-to-human transmission.

Climate change passes both #1 and #2. There is known, first-principles mechanism of action that you can dig into like into a fractal, if you want to explore consequences for various aspects of the ecology and society. There is a known mechanism of action that makes it into an actual threat - food shortages, mass migrations, political unrest, leading to wars, starvation, suffering and death. We know how fragile our civilization is (present COVID-19 situation makes a lot of weak points clearly visible). There's lot of historical data supporting the above and the threat evaluation - for instance, we're seeing how increasing average temperatures are impacting agriculture and habitability of some regions on Earth.

One more clarification: the reason climate change is dangerous is not because it threatens plants and animals. It's because it threatens human civilization. We care about nature not because it is, but primarily because we need it to live. Moreover, since almost no one would enjoy suddenly being thrown back to preindustrial age on a thoroughly broken planet, we can safely say that the concern is about survival of existing civilization, not survival of human species (not to mention, you can't fix the damage without technology, and you won't have technology when civilization collapses). Human civilization is very fragile.

Also: just saying "tail risks" isn't meaningful. You have to actually estimate the sum of the tail to determine whether it's even worth paying attention to. Which circles us back to heuristics #1 and #2, which - in my opinion - say that microplastics are a secondary issue to climate change.

[tomp]

Good answer, thanks!